1. Field of the Invention
This invention relates generally to purchasing and reservation systems and, in particular, the present invention relates to improvement of yield management with respect to perishable commodities such as airline seats, hotel rooms and the like.
2. Description of the Related Art
Common carriers such as commercial buses, trains, and airlines, and service industries such as hotels and rental car companies, face complex issues when conducting strategic and operational planning. Businesses of this type deal with “perishable commodities” which are defined as commodities that cannot be inventoried and share three common characteristics: perishability, “fixed” capacity, and segmentability. Perishability means that each commodity ages or becomes unavailable, and thus has no value, after a certain date, time or similar temporal event. “Fixed” capacity implies a high cost of adding an incremental unit such that capacity is regarded as static and unchanging. Segmentability refers to the ability to segment customers based on a willingness to pay using different rates and/or different purchase restrictions, such as the date of purchase relative to the date of use. Examples of perishable resources include airline seats, hotel room nights, rental car days and similar products or services such as described in L. R. Weatherford & S. E. Bodily, A Taxonomy and Research Overview of Perishable-Asset Revenue Management Yield Management, Overbooking, and Pricing, 40 Operations Research 5, pp. 831-44 (1992), the disclosure of which is incorporated herein by reference.
Organizations marketing and selling perishable commodities spend numerous hours trying to choreograph the interrelated elements of scheduling, routing, and crew/staff rotations while maximizing profits and efficiency. Maximum profits are achieved when all of the available perishable commodities (e.g., with respect to airlines, all seats on a given flight) are sold on the perishing date (e.g., with respect to airlines, at the time the given flight departs). Maximum customer satisfaction occurs when perishable commodities reserved by consumers are available on the perishing date. The marketer/seller of perishable commodities must therefore constantly balance these two competing interests so that all of the commodities are sold and are available for all those who reserved them.
The terms “revenue management” and “yield management” are now common terms in service industry parlance to describe the use of statistical analysis to manage itinerary control, inventory control, over-booking and pricing so as to increase the revenue yield per unit of available capacity. Based on the statistical analysis, forecasting, optimization models, and the like, determinations are made as to which reservation requests to accept and which to reject in order to maximize revenues.
The airline industry presents a typical example of a service industry which utilizes yield management techniques to try to maximize profits while coping with the complicated operational issues inherent to the industry. An airline passenger may have five or more carriers to choose from when planning a trip from point A to point B. Airlines are constantly seeking ways to maximize their efficiency and profits and make tremendous efforts to win and maintain customer loyalty. Airline passengers' primary criteria in selecting airlines include safety, comfort, and timeliness. Given the relatively wide selection of carriers available to airline passengers, and the ease with which reservations can be made and changed, a failure in any of these areas by an airline is likely to result in a migration of passengers to other airlines.
It is a well-known practice in the airline industry to overbook flights in an attempt to assure that the flights are fully loaded with passengers, thereby maximizing the profits for the airlines. The policy of overbooking is based upon practical considerations. For various reasons, not all flights reserved are actually purchased, i.e., while they may have been reserved, they do not actually “materialize”. For example, air travelers frequently reserve seats which they have no intention of using so that they may be assured of having the most convenient itinerary possible. Take, for example, an air passenger who is based in Philadelphia and needs to be in London, England on a particular Tuesday by 1 p.m. The passenger might reserve a first itinerary comprising a direct flight from Philadelphia to London on British Airways, arriving in London at 12:30 p.m. on Tuesday; a second itinerary comprising a first leg from Philadelphia to John F. Kennedy Airport in New York on American Airlines and a second leg from John F. Kennedy Airport to London via British Airways, arriving in London at 11:30 a.m. on the same Tuesday; and a third itinerary comprising a first leg from Baltimore/Washington International Airport to Dulles Airport in Washington, D.C., and a second leg from Dulles Airport to London, arriving in London, again, on the same Tuesday at 10:00 a.m., both legs being on U.S. Airways. Clearly, the passenger can only use one of the itineraries; however, for convenience sake, the user may wait until the last minute to decide which of the three itineraries to utilize, and the user may or may not proactively cancel the two unused (i.e., unmaterialized) itineraries.
Similarly, the passenger may not know with any certainty what time he or she will be able to leave London on a return flight. For example, suppose that the passenger is traveling to London on business and will be a conducting meeting having an unknown duration. The passenger may reserve several outgoing flights spaced several hours apart so as to be assured of having a reservation on a flight leaving within a reasonably short time after the conclusion of the business meeting.
While providing convenience for the air passenger, such reservation practices make it particularly difficult for airlines to assure that all flights depart without empty seats. To compensate for unmaterialized reservations, airlines have adopted the policy of overbooking flights with the understanding that a certain percentage of the seats on “reserved” status by passengers will never actually materialize. In a perfect world, the airlines could always tell with precision precisely how many passengers would over-reserve for a particular flight and would then overbook for that flight by the exact number so that all seats would be filled. In reality, however, it is impossible to predict precisely how may reservations will not materialize; thus, airlines frequently end up with either too few seats sold, thereby losing revenues by flying aircraft with empty seats, or too many seats sold, requiring the airlines to “bump” passengers onto the next available flight to their destination. While most airlines will in some manner compensate passengers that have been bumped, for example, by providing them with vouchers good towards future flights on the airline, free hotel accommodations, and the like, such a practice, is costly for the airlines, is usually extremely inconvenient to the airline traveler, and can lead to once-loyal passengers migrating to a competitor airline.
In an attempt to overcome the above problems, systems have been developed which track the frequency with which a particular flight experiences overbooking or underbooking and, based on this statistical analysis, increases the point at which that particular flight is considered “closed” to a number greater than 100% of the capacity of the aircraft, with the exact percentage greater than 100% being based upon the historical data for that flight. Examples of such systems can be found in, for example, U.S. Pat. No. 5,918,209 to Campbell et al., U.S. Pat. No. 5,255,184 to Hornick et al., and U.S. Pat. No. 4,775,936 to Jung, all of which are incorporated fully herein by reference.
Each of the prior art systems known to the applicant involve statistical analysis of the perishable commodity in question (e.g., the particular flight, airline seat, hotel room, rental car) to determine the history of booking with respect to the perishable commodity over a period of time. Thus, for example, a particular flight (e.g., Flight Number 250 from Philadelphia to London) and/or combination of legs comprising an entire itnerary, might be analyzed to determine the likelihood that the particular flight(s) will be fully sold out, based on the overall past history of overbooking or underbooking for the flight(s). Additional factors considered by the prior art systems may include whether or not a particular event is associated with the flight (e.g., did a particular flight experience different sales characteristics when the flight was associated with travel to and from the Olympic Games?), or whether any sales promotions are associated with the flight (e.g., was it necessary to offer discounted fares in order to fill up the seats?). While the use of such systems provides assistance to the airlines, their focus is always on very general statistical history of a specific booking itself, and guesses regarding the impact of outside factors such as event association or promotional fare structures.
None of the prior art systems analyze: details of the individual reservations so that the reservations can be characterized as having one or more traits; characteristics of the consumers who reserve the bookings so that the consumers (or potential consumers) can be characterized as having one or more traits; or the reasons why the person making the reservation actually did (or did not) purchase the reserved booking. The Applicant has determined, however, that it is only by understanding why a particular booking actually materializes that better prediction models can be developed so as to optimize the yield management or revenue management system. If the details of the booking (other than simply the flight number) and characteristics of the persons making the reservations were to be factored into the analysis, the effectiveness of the overbooking policy by the airlines could be increased. However, none of the prior art systems attempt to make such an analysis.
Accordingly, it would be desirable to have a method and system for gathering and storing information about prior reservations, details of the perishable commodity reserved, and characteristics related to the person making the reservation, and using this information to determine, with more precision than is available in the prior art, the likelihood that current pending reservations will materialize.